Factors Determining Differences in Soil pH in Adjacent Conifer and Alder-Conifer StandsDan Binkley* and Phillip Sollins

ABSTRACTTree species may differ in their influence on biogeochemical cycles,

leading to differing rates of soil acidification. Over time, quantitativeand qualitative changes develop in the characteristics of the soilexchange complex. Three such characteristics regulate soil pH: (i)the quantity of acids present, which can be represented as the totalcation-exchange capacity (CEC); (ii) the degree of dissociation ofthe acids, commonly called base saturation; and (iii) the affinity ofthe acids for H+, or acid strength, which represents the compositepA. (negative log of the acid ionization constant) of the exchangecomplex. We examined the importance of these three factors in ex-plaining the differences in soil pH between adjacent stands of con-ifers [primarily Douglas fir, Pseudotsuga menziesii (Mirbel) Franco]and conifers mixed with N-fixing red alder (Alnus rubra Bong.). Ata low-productivity site (Wind River, WA), the pH of 0 to 0.15 m ofsoil from both alder-conifer and conifer stands averaged 4.3 in 0.01M CaCl2. The pH values were the same, however, only becausehigher base saturation in the alder-conifer stand was offset bygreater acid strength. At a more productive site (Cascade Head, OR),soil pH (in 0.01 M CaCl2) averaged 3.7 in the alder-conifer standbut 4.4 in the conifer stand. The difference in pH resulted primarilyfrom greater acid strength of soil organic matter under alder, andsecondarily from lower base saturation of the exchange complex.These results underscore the importance of considering qualitativechanges in soil organic matter as factors driving changes in soil pHand other parameters.

DECLINES IN SOIL pH under red alder, a N-fixingtree species common in the Pacific Northwest,

have been noted for more than two decades. The mag-nitudes of the declines have varied, ranging from es-sentially nil (e.g., Binkley, 1983; Binkley et al., 1984)to one unit in 50 yr (e.g., Franklin et al., 1968; Bor-mann and DeBell, 1981; Van Miegroet and Cole, 1984;Van Miegroet et al., 1989). Van Miegroet and Cole(1984) attributed the decrease in soil pH to decreasedbase saturation. They suggested that the lower basesaturation resulted from the production of H+ in ni-trification, with FT displacing base cations, which thenleached from the soil with NO3.

Base saturation, however, is only one of three factorsof the soil exchange complex that determine soil pH.Binkley et al. (1989) noted that the pH of a solutionis determined by the quantity of weak acids (both pro-tonated and dissociated), the degree of dissociation ofthe acids, and the strength of the acids (the pATJ. Insoils, stabilized weak acids are composed of organicmatter and crystalline and amorphous clays (Truog,1945; Perdue, 1985). The quantity of acid can be rep-resented by the total CEC of the soil. The degree ofdissociation of the exchange complex is gauged by baseD. Binkley, Dep. of Forest and Wood Sciences, Colorado StateUniv., Ft. Collins, CO 80523; and P. Sollins, Dep. of Forest Science,Oregon State Univ., Coryallis, OR 97331. This project was fundedprimarily by National Science Foundation Grant BSR-841678, andalso by the Integrated Forest Study of the Electric Power ResearchInstitute through Oak Ridge National Laboratory. Received 16 Nov.1989. "Corresponding author.

Published in Soil Sci. Soc. Am. J. 54:1427-1433 (1990).

saturation (proportion of the exchange sites occupiedby so-called base cations). Although cations are notbases in any chemical sense (Leeper, 1948), base cat-ions represent the dissociated portion of the exchangecomplex, just as the undissociated portion is repre-sented by H+ and Al. Acid strength in soils derivesfrom the composite contribution of many types ofacids of varying p.Ka.

The relative importance of these three factors canbe examined with empirical titration curves (Binkleyet al., 1989). The total quantity of acids present in thesoil is represented by the quantity of acid or baseneeded to lower or raise the pH of the suspension toa defined end point. The quantity of acid required toreach an end point is defined as the acid-neutralizingcapacity (ANC), and the quantity of base required isdefined similarly as the base-neutralizing capacity(BNC; Stumm and Morgan, 1981). The degree of dis-sociation of the soil acids can be represented by thesum of the positive charge contributed by K, Na, Ca,and Mg. The cumulative acid strength is described bythe shape of the titration curve. As described below,sensitivity analysis can be used to judge the contri-bution from each factor to the difference in pH be-tween two soils.

In this project, we applied this empirical model tosoil samples from two pairs of conifer and alder-con-ifer stands to determine the relative importance ofeach factor in determining soil pH.

METHODSSite Description

The study sites are described in detail in Tarrant (1961)and Franklin et al. (1968). Briefly, the low-productivity WindRiver site in western Washington has a site index for Douglasfir (without alder) of 25 m at 50 yr. The ecosystem N capital(biomass + soil to 0.9 m) was about 153 kmol/ha (2140 kg/ha) in a pure Douglas fir stand and about 353 kmol/ha (4940kg/ha) in an adjacent alder-Douglas fir stand. Nitrogen min-eralization in resin cores (Binkley and Hart, 1989) was al-most undetectable in the Douglas fir stand, but averaged 12.5kmol ha"1 yr1 (175 kg ha-1 yr1) annually in the alder-Doug-las fir stand. Only about 0.03 kmol ha-1 yr1 of NO3 (0.4 kgN ha-1 yr1) leached from the conifer stand (based on ten-sionless lysimeters at 0.8 m and an extrapolated site waterbalance), and 0.18 kmol/ha (2.5 kg N ha-1 yr1) from thealder-conifer stand. The annual rate of accumulation of K+ Ca + Mg in biomass was 0.45 kmolc/ha for the coniferstand, and 0.67 kmolc/ha for the alder-conifer stand. Thesoil, an unclassified Andic Haplumbrept, has a silty clayloam texture, and developed in Tertiary andesitic or rhyoliticparent materials with some pumice and basaltic gravel ofPleistocene origin (Miller and Tarrant, 1983). The site ex-perienced repeated wildfires before plantation establishmentin 1929. The alder were planted in a 22-m-wide strip througha 1000-m portion of the Douglas fir plantation.

The Cascade Head study site on the coast of Oregon isvery productive, with a site index for Douglas fir (withoutalder) of 40 m at 50 yr. In 1935, a 0.2-ha conifer plot wasestablished by cutting all alders in an 8-yr-old, naturally re-generated stand containing red alder, Douglas fir, western

1427

1428 SOIL SCI. SOC. AM. J., VOL. 54, SEPTEMBER-OCTOBER 1990

hemlock [Tsuga heterophylla (Raf.) Sarg.], and Sitka spruce[Picea sitchensis (Bong.) Carriere]. The alders were retainedin an adjacent 0.2-ha plot. In 1985, the total ecosystem Ncapital of the conifer stand was 747 kmol/ha (10 460 kg/ha),compared with 1033 kmol/ha (14460 kg/ha) in the alder-conifer stand (Binkley et al, 1989, unpublished data). Nitro-gen mineralization at 0 to 0.3 m was estimated to be about2.9 kmol ha-1 yr1 (41 kg ha-1 yr1) in the conifer stand,compared with 12.5 kmol ha-1 yr1 (175 kg N ha-1 yr1) inthe alder-conifer stand. Nitrate leaching was about 0.5 kmolha-1 yr1 (7 kg N ha-1 yr1) from the conifer stand, and 2.3kmol ha-1 yr1 (32 kg N ha-1 yr1) from the alder-coniferstand. The average annual rate of accumulation of K + Ca+ Mg in biomass was 0.70 kmoLyha for the conifer standand 0.55 kmoljha for the alder-conifer stand. The soil is awell-drained Typic Dystrandept of the Astoria silty clay loamseries, developed in tuffaceous siltstone over basalt.

Soil Sampling and AnalysisTen soil pits were dug in each stand, located at random

distances (on alternate sides) from random points on a tran-sect through the middle of each plot. The pits were approx-imately 0.5 by 1.0 m, and >1 m deep. Soil samples forchemical analysis were taken from 0 to 0.15, 0.15 to 0.40,0.40 to 0.65, and 0.65 to 0.90 m at Wind River, and from0 to 0.15, 0.15 to 0.30, 0.30 to 0.60, and 0.60 to 0.90 m atCascade Head. These depths corresponded roughly to theAh, Bl, B2, and Cl horizons in all stands. The volume oflarge rocks was estimated by weighing all rocks removedfrom each pit, with conversion to volume based on the den-sity of a subset of rocks. Bulk density of the rock-free soilwas determined by weighing samples of known volumetaken from the walls at each depth interval in each pit.

Soil C was analyzed with a LECO 12 C analyzer (LECOCorp., St. Joseph, MI). Extractable soil cations were deter-mined by atomic absorption spectrophotometry on 100-mLextracts (1 M NaCl) of 10-g samples. Extractable Al (whichmay be underestimated with extraction with NaCl) was as-sumed to have a charge of +3. Total K, Ca, and Mg weredetermined by inductively coupled plasma spectrophoto-metry on samples digested with HF, H2SO4, and HC1O4 (Limand Jackson, 1982, Method 1-6).

Five methods were used to characterize soil acidity. SoilpH was measured in solutions of deionized water, 0.01 MCaCl2 and 1 AfKCl, with soil weight/solution volume of 1:5.The effect of ambient differences in ionic strength on pHwas determined by comparing the pH of samples in deion-ized water and in 0.01 M CaCl2, and by analyzing the anionspresent in water extracts (50 mL of deionized water and 5g of dry soil, analyzed with a DIONEX anion chromatograph[DIONEX Corp., Sunny vale, Ca]).

Titration curves were used to estimate the ANC and BNCof the soils. Three composite soil samples were preparedfrom the original 10 air-dry samples from each of the fourhorizons in each stand (resource limitations prevented usingall 40 original samples per stand). A set of 50-g subsampleswere placed into 125-mL specimen cups, and 50 mL of 1 MKC1 were added to buffer the ionic strength of the suspen-sion. This set received consecutive additions (in 0.5-mL in-crements) of 0.01 MHC1, for poorly buffered soils, or 0.05M HC1 for more highly buffered soils. The pH of the su-pernatant remained relatively stable between 12 and 36 hafter addition of the acid or base; we, therefore, measuredpH between 20 and 28 h after each addition. Acid additionswere stopped when pH 3.0 was reached. A second set ofsubsamples received similar increments of 0.01 or 0.05 MNaOH, stopping when the pH passed 8.2.

The empirical method of Binkley et al. (1989) was usedto examine the relative importance of mechanisms (acidquantity, acid dissociation, and acid strength) in determining

the difference in pH between conifer and alder-conifer soils.The importance of differences in the quantity of acids presentin the soil was evaluated by taking the titration curve forone soil, holding the content of K + Ca + Mg constant,and stretching the curve to match the greater ANC + BNCof another soil. The equation used to derive the adjustedcurve was:

(ANC)new = (ANC)original[ANC + BNC]ori;

[ANC + BNC]n[1]

where (ANC)ncw gives the new x coordinates for each y co-ordinate on the shifted curve, (equivalent to the concentra-tion of base cations that balance the charge); (ANC)original isthe base-cation concentration for the original curve, and[ANC + BNC]new and [ANC + BNC]original are the totalquantities of acid present in the new soil and in the originalsoil. The pH of the new curve at the point of zero additionof acid or base equals the expected pH of the new soil if theANC + BNC were shifted to match that of another soil (seeBinkley et al., 1989, for graphical illustrations of this em-pirical method).

The importance of changes in the dissociation of the soilacids (equal to changing the content of base cations) wasassessed by adding or removing a quantity of OH from thetitration curve to match the desired change in dissociationof the exchange complex.

The importance of changes in acid strength was deter-mined by adjusting the titration curves for the alder-conifersoil with the above equation to match the acid quantity ofthe conifer soil, and then adjusting the dissociation of theexchange complex (OH added) to match the conifer soil. Theresidual difference in pH after these adjustments representsthe importance of the difference in acid strength.

Analyses of variance, used to examine differences inmeans, were calculated with SYSTAT (SYSTAT, 1985) witha probability of a Type I error set at 0.10.

RESULTS AND DISCUSSIONCharacteristics of the Soil Exchange Complex

For soil samples from Wind River, pH in water wassignificantly lower (at P < 0.10) in the alder-coniferthan in the conifer soils. For pH in salt solutions, theonly significant difference was a greater pH in thealder-conifer soils at 0.40 to 0.90 m (Table 1). Acid-neutralizing capacity of the alder-conifer soils ex-ceeded that of the conifer soils only at 0 to 0.15 m(Table 1, Fig. 1). Base-neutralizing capacity at pH 5.5(BNCpH 5.5) was about 1.35 times higher in the conifersoils at 0!65- to 0.90-m depth than in the correspond-ing alder-conifer soils. The BNCpH 8.2 was higher inthe alder-conifer soil at 0 to 0.15 m, but lower at 0.40to 0.60 m. The buffer-intensity distribution (change inpH per unit of OH added as a function of pH; Stummand Morgan, 1981, p. 160) showed that the alder-con-ifer soils were buffered better at any pH with respectto additions of OH than the conifer soils, but less buff-ered against additions of H+ (Fig. 2).

Concentrations of extractable cations were generallyhigher at 0 to 0.40 m in the alder-conifer soils at WindRiver than in the conifer stand (Table 2). Althoughno differences were significant for the lower depths,the apparent decrease in extractable cations at 0.40 to0.90 m in the alder soils roughly balanced the increasesat 0 to 0.40 m (Table 3). The content of extractablebase cations was significantly greater in the alder-con-

BINKLEY & SOLLINS: pH DIFFERENCES IN ADJACENT TREE STANDS 1429

ifer soils at 0 to 0.40 m, but significantly smaller atthe 0.40- to 0.65-m depth. The net difference in ex-

tractable K + Ca + Mg was only 9 kmolg/ha at the0-to 0.90-m depth, which is too slight to be statistically

Table 1. Means of soil pH and acid- and base-neutralizing capacities (ANC and BNC). Standard deviations are given in parentheses; n10/stand.

BNC

pH ANC to pH 3 topHS to pH 8.2

Location,stand depth

m

Wind RiverConifer

0-0.150.15-0.400.40-0.650,65-0.90

Water

5.4 (O.l)t5.3 (O.l)f5.2 (O.l)t5.2 (O.l)f

0.01 M CaCl2

4.3 (0.1)4.3 (0.1)4.1 (0.1)4.1 (0.1)

1MKC1

4.0 (0.0)4.1 (0.1)3.8 (O.l)f3.6 (O.l)f

mmol H/kg

259 (31)t317 (36)t187 (12)141 (17)

kmolH/ha

201563360263

mmol OH/kg

62(9)66 (14)

167 (53)218 (30)f

kmolOH/ha

48117321407

mmol OH/kg

283 (29)t248 (17)329 (39)f361 (34)f

kmolOH/ha

220440633673

Sum

Alder-conifer0-0.150.15-0.400.40-0.650.65-0.90

Sum

Cascade HeadConifer

0-0.150.15-0.300.30-0.600.60-0.90

SumAlder-conifer0-0.150.15-0.300.30-0.600.60-0.90Sum

5.1 (0.3)t5.1 (0.2)f5.0 (0.2)f4.9 (0.2)f

5.4 (0.4)t5.4 (0.2)f5.2 (0.2)t4.9 (0.2)f

4.3 (0.4)t4.8 (0.4)f4.8 (0.3)f4.7 (0.2)f

4.3 (0.2)4.2 (0.1)4.1 (0.1)4.0 (0.1)

4.4 (0.4)t4.4 (0.2)f4.3 (0.2)4.1 (0.1)

3.7 (0.3)t4.1 (O.l)f4.2 (0.1)4.1 (0.1)

4.0 (0.1)4.1 (0.1)3.9 (O.l)t3.7 (O.O)f

4.0 (0.2)4.1 (0.1)4.0 (0.0)3.9 (0.0)

3.6 (0.3)4.0 (0.0)4.0 (0.0)3.9 (0.0)

220 (31)f270 (34)f200 (33)161 (16)

429 (60)t736 (95)t671 (39)384 (25)

227594 (36)f665 (26)415 (69)

1387

150468385311

1314

270545

1026578

2419

180440

1 125735

2480

87 (24)61(8)111 (23)159 (16)f

167 (48)144 (32)126 (17)189 (22)

298(115)181 (28)137 (1)168 (20)

893

59106214308687

105107193285690

194134232297857

377 (69)t247 (20)264 (10)t306 (16)t

686 (50)f592 (24)t520 (27)447 (22)

825 (97)t689 (52)t537 (27)430 (20)

1966

257428508592

1785

433439796673

2341

537510907761

2715t Means differ at P < 0.10 between stands at each depth and location.

-0.3-0.10.1 0.3 0.5 0.7-0.3-0.10.1 0.3 0.5 0.7 0.9 -0.3-0.10.1 0.3 0.5 0.7-0.3-0.10.1 0.3 0.5 0.7 0.9OH"addedmo^/kg

Fig. 1. Titration curves showing pH of supernatant 20 to 28 hr after addition of HC1 or NaOH to soils from four depths. A = Wind River,B = Cascade Head; dashed line = conifer soil, solid line = alder-conifer soil.

1430 SOIL SCI. SOC. AM. J., VOL. 54, SEPTEMBER-OCTOBER 1990

Table 2. Concentrations of total C and exchangeable (Exch.) cations and base saturation. Standard deviations are given in parentheses; n =10/stand.

Location,stand, depth Total C

m mmol/kgWind River

Conifer0-0.15 3650(300)f0.15-0.40 1 290 (90)t0.40-0.65 450 (50)f0.60-0.90 230 (30)f

Alder-conifer0-0.15 m 5 580 (750)t0.15-0.40 1690(160)f0.40-0.65 660 (80)t0.60-0.90 420 (40)f

Conifer0-0.15 9810(730)0.15-0.30 6 400 (470)0.30-0.60 3 870 (360)0.60-0.90 1 760 (160)

Alder-conifer0-0.15 9 850 (640)0.15-0.30 6930(300)0.30-0.60 4 080 (400)0.60-0.90 1 730 (180)

Ca

Total

Mg

Exch. Total

890 8.3 (0.5)f 790635 2.5 (0.5)t 770345 13.0 (0.6) 740215 15.2 (0.6) 725

915 22.5 (0.7)t 725725 6.6 (0.7)f 740515 6.4 (0.8) 770400 12.7 (0.8) 770

120 12.1 (O.l)t 510100 5.5 (0.1) 59095 2.3 (0.2)t 62565 2.1 (0.2) 690

115 6.2(0.3) 500100 5.7 (0.3) 53085 1.4 (0.4) 62080 1.6 (0.4) 670

K

Exch. Total Exch.

3.2 (0.6)t 230 3.7 (0.4)1.9 (0.6)t 225 3.2 (0.3)9.9 (1.2) 174 2.3 (0.5)

11.5(1.3) 130 1.7(0.3)

8.1 (1.7)t 215 3.4 (0.3)5.9 (0.9)f 225 3.1 (0.1)6.6 (1.2) 215 1.7 (0.2)

10.4 (2.5) 170 1.2 (0.0)

19.7 (6.7)t 130 2.0 (0.2)11.8(5.0) 150 1.1(0.2)6.4 (2.1)t 160 0.6 (0.1)4.8 (0.7) 185 0.6 (0.1)

4.1 (0.8) 130 2.1 (0.5)5.9 (3.4) 145 3.2 (0.3)2.2(1.2) 170 0.7(0.1)2.7 (0.9) 185 0.7 (0.2)

Al

Exch.

32.4 (7.3)28.0 (7.1)f83.0 (29.9)

150.2 (18.5)

41.9 (5.0)45.5 (3.3)80.9 (10.0)

124.3 (22.7)

88.6 (6.0)f50.6 (15.7)52.9(11.6)

118.2(27.5)

138.0 (3.7)f75.9 (17.4)48.6 (9.0)61.1 (19.3)

Basesaturation

0.32f0.22f0.24f0.16

0.45f0.26f0.16f0.17

0.29f0.29f0.1 7f0.07

0.09f0.1 8f0.09f0.08

t Means differ at P < 0.10 between stands at each depth and location.

Table 3. Total and exchangeable (Exch.) soil element pools at the 0- to 0.9-m depth.

CaLocation, stand Total C Total Exch.

knWind River

lol/ha

Conifer 6490 21280 641Alder-conifer 8 825 31 080 636

Cascade HeadConifer 19495 40200 184Alder-conifer 21 580 43 900 126

20

12

0 4

1°

£20

12

40 . . . . . . . . . . - . . . . . . . . . .

A

A 0-0.15 m >

/' 0.40-0.65 m

20 •

J "'•Jt 0.1 5-0.40 m .^wt ^ '

sf 20

///0.65-0.90 m

•y 4'

Mg

Total Exch.

38 400 45839 200 481

28 000 37729 650 147

B

0-0. 15m

0.30-0.60 m

K Al

Total Exch. Exch.

8 950 163 5 15010 400 136 5 050

7 400 41 3 5307 970 45 3 210

0.1 5-0.30 m

0.60-0.90 m

3 5 7 3 5 7 9 3 5 7 3 5 7 9PH

Fig. 2. Buffer-intensity distribution, representing change in pH/OH as a function of pH. A high value indicates low buffer intensity. A =Wind River, B = Cascade Head; dashed line = conifer soil, solid line = alder-conifer soil.

or ecologically significant. The extractable pools com-posed a small fraction of the total pool of K, Ca, andMg (Table 3).

For soil samples from Cascade Head, pH in water

was significantly lower at all depths in the alder-con-ifer than in the conifer soils (Table 1). The alder-con-ifer soils also showed lower pH in dilute salt solutionsat 0- to 0.30-m depth, but not in 1 M KC1; ANC and

BINKLEY & SOLLINS: pH DIFFERENCES IN ADJACENT TREE STANDS 1431

Table 4. Water-soluble anions, H* concentrations, and pH for three composited samples for 0- to 0.15-m-depth soil from each stand. Standarddeviations are given in parentheses.

Location, stand pH Cl NO3 SO4 Alkalinity Sum of anions- ftmoL/L -

Wind RiverConiferAlder-conifer

Cascade HeadConiferAlder-conifer

5.4 (0.1)5.1 (O.l)t

5.3 (0.1)4.6 (0.3)f

4.3 (0.8)t7.9 (0.9)t

4.6 (0.8)T24.0 (9.1)f

56(5)55 (13)

66 (12)50 (20)

10 (6)t185 (12)f

100 (9)f350 (55)f

22 (2)t13 (3)f

25 (3)t19 (2)t

72 (16)63 (46)

53 (l)t11 (15)f

175 (10)f330 (45)f

270 (10)f450 (20)f

t Means differ at P < 0.10 between stands at each depth and location.

Table 5. Measured and expected pH in 1 M KC1 in the conifer stands if single factors were set equal to the value for the alder-conifer stands.

Measured pH Expected pH in conifer soil

Location Depth, m Alder/conifer Conifer ANC + BNCf Base cations Acid strengthWind River

Cascade Head

0-0.150.15-0.400.40-0.650.65-0.900-0.150.15-0.300.30-0.600.60-0.90

4.04.13.93.73.64.04.03.9

4.04.13.83.63.94.14.03.9

4.04.23.83.64.04.14.13.9

4.34.03.83.63.84.14.03.9

3.94.03.83.73.74.04.03.9

t Acid-neutralizing capacity plus base-neutralizing capacity.

BNC values differed significantly between stands at alldepths (Table 1, Fig. 1). The buffer-intensity distri-bution differed markedly between stands (Fig. 2), es-pecially at the 0- to 0.15-m depth, where the conifersoils were buffered better against change in pH nearthe ambient pH, but more poorly buffered betweenpH 5 and 7.

Contents of extractable K + Ca + Mg at CascadeHead were lower at the 0- to 0.15-m depth in the alder-conifer stand than in the conifer stand (Table 2), andthe base saturation was also lower at the 0- to 0.60-mdepth. In contrast to the Wind River site, the differ-ences in K + Ca + Mg appeared to result from re-moval from the profile rather than from redistributionamong horizons. The difference of 284 kmolc/ha (0-0.90-m depth) represents almost one-half of the ex-tractable K + Ca + Mg present in the conifer soil. Ifthe alder-conifer soil originally had the K + Ca +Mg content of the conifer soil, the average annual rateof removal for 55 yr would have been about 5.2 kmol,/ha. The apparent depletion of extractable cations rep-resents a very small fraction (<0.05) of the totalamount in the soil (Table 3).

Factors Determining Soil pHSoil pH in water was lower under the alder-conifer

stand at both locations, caused in part by the greaterionic strength in the water suspensions. The greaterionic strength resulted primarily from the greater con-centrations of NO3 in the alder-conifer soils (Table4). For example, the concentration of water-solubleanions from the alder-conifer stand (0-0.15 m) atWind River exceeded that of the conifer stand byabout 160 /umolc/L. Of this extra charge, H+ balancedonly 3.6 nmolJL, but this difference in H+ concentra-tions was sufficient to drop solution pH by 0.3 units.The difference in anion concentrations between stands

at Cascade Head was about 180 jumolc/L, and H+ bal-anced about 20 /umolc/L of the charge and lowered pHby 0.6 units. These patterns are an example of thetendency of increased ionic strength to lower solutionpH in soils with significant exchangeable acidity (seeReuss and Johnson, 1986; Richter et al., 1988), with-out any substantial change in the exchange complex.

The evaluation of the relative importance of acidquantity, dissociation, and strength can be illustratedwith data for the 0.15- to 0.40-m depth from the WindRiver stands. Within this depth interval, the conifersoil contained 7.5 mmolc/kg of K + Ca + Mg (Table2), with ANC + BNC (between pH 3 and 5.5) of 383mmolc/kg (Table 3). The acid quantity for the 0.15-to 0.40-m soil from the alder-conifer stand at WindRiver was 331 mmolc/kg. If the acid quantity of theconifer soil was reduced to match that of the alder-conifer soil (a 14% reduction), the resulting increasein based saturation would increase soil pH from theobserved 4.1 to 4.2 (Table 5). This rise was calculatedby proportionally shortening the conifer soil's titrationcurve by 14% at each point, and then reading the pHat the level of zero addition of acid or base.

The 0.15- to 0.40-m soil samples from the WindRiver conifer stand contained 7.5 mmolc/kg of K +Ca + Mg, compared with 15.6 mmolc/kg for the alder-conifer soil. The addition of 8.1 mmolc/kg of K + Ca+ Mg to the conifer soil would be equivalent to ti-trating the soil with 8.1 mmolc/kg of OH; therefore,the predicted pH can be read from the original titrationcurve by moving up the curve a distance correspond-ing to the 8.1 mmolc/kg of OH added. The pH of theconifer soil that would be expected if it contained thequantity of base cations found in the alder-conifer soilwould be essentially unchanged (<0.1 units).

The importance of differences in acid strength wereevaluated by adjusting the curve (samples from 0.15to 0.40 m in the alder-conifer stand) to have the ANC

1432 SOIL SCI. SOC. AM. J., VOL. 54, SEPTEMBER-OCTOBER 1990

+ BNC and the content of K + Ca + Mg of the soilsfrom the conifer stand. After these adjustments, thepH at the zero point of addition of acid or base rep-resents the pH expected for the conifer soil if it hadthe acid strength of the alder-conifer soil. For thisdepth, differences in acid strength appeared to be un-important (Table 5).

At Wind River in general, acid quantity did notdiffer enough between stands to be important, but dis-sociation of the exchange complex (base saturation)and acid strength differed substantially in the uppersoil (Table 5). If the 0- to 0.15-m conifer soil had thehigher concentrations of K + Ca + Mg found in thealder-conifer stand, the expected pH in KC1 wouldhave been 4.3 rather than the observed 4.0. If the con-ifer soil were altered to have the acid strength of thealder-conifer soil (but retaining the original ANC +BNC and K + Ca + Mg), the expected pH (0-0.15m) would have been only 3.9.

The most important difference in acidity factors be-tween soil from the Cascade Head stands was thegreater acid strength under the alder-conifer stand(Table 5). If only the acid strength of the conifer soilwere changed to match the alder-conifer soil, the pHwould have been 3.7 rather than the observed 3.9.Changing the concentrations of base cations alonewould have lowered pH to 3.8. The difference in acidquantity was not important. The stronger acidity andlower concentrations of base cations in the alder-con-ifer stand combined to produce the observed pH of3.6. In deeper horizons, pH differed little betweenstands, thus all factors were of marginal importance(Table 5).

Our results indicate that the hypothesis that de-creased base saturation under alder is driven by H+

production during nitrification may be correct in somecases, but it may not be a complete explanation. Thelack of difference in soil pH at Wind River, wherenitrification was roughly balanced by NO3 assimilation(as gauged by the low rate of NO3 leaching), appearsconsistent with the hypothesis. However, our resultsindicate that soil pH at 0 to 0.30 m would have beenlower in the alder-conifer stand if the increase in acidstrength had not been offset by an increase in basesaturation. At Cascade Head, the low base saturationin soils from the alder-conifer stand relative to thosefrom the conifer stand was consistent with the hy-pothesis, but the large difference in base saturationappeared less important than the greater acid strengthof the exchange complex. The importance of differ-ences in acid strength in controlling pH was also dem-onstrated by Binkley and Valentine (1990), whoreported that greater acid strength accounted for themajority of difference in pH between soils under Nor-way spruce [Picea abies (L.) Karsten], white pine (Pi-nus strobus L.) and green ash (Fraxinus pennsylvanicaMarshall). Binkley et al. (1989) found that a drop inacid strength partially offset the 20-yr decline in soilpH in a loblolly pine (Pinus taeda L.) stand that ac-companied a decrease in base saturation (driven byH+ production associated with accumulation of Ca,Mg, and K in the biomass). The important role playedby changes in the strength of soil acids across time andunder the influence of plant species underscores the

importance of considering changes in soil acid strengthin long-term assessments of soil acidification.

The low pH of the upper soil in the alder-coniferstand at Cascade Head relative to the conifer standmay result in substantially higher concentrations of Alin soil solutions. The low base saturation would makethe soil-solution concentrations of Al especiallyresponsive to changes in ionic strength (Reuss andJohnson, 1986; Reuss, 1989). Further, a reduction ofone pH unit from 5.4 to 4.4 might increase 1000-foldthe concentration of dissolved Al in equilibrium withAl minerals (Lindsay, 1979). These implications war-rant direct assessment of the effects of alders on Albiogeochemistry.

ACKNOWLEDGMENTSThe field work and laboratory analyses were provided

cheerfully by Randy Bell, Don Sachs, Amy Home, CarolGlassman, Anne Russell, and Ute Valentine. Dave Valentineassisted in data analysis. This paper is a contribution of theH.J. Andrews Ecosystem Research Group.

SHEPARD & MITCHELL: LONGEVITY OF POTASSIUM FERTILIZATION RESPONSE 1433

ERRATUM

Factors Determining Differences in Soil pH in Adjacent Conifer and Alder-Conifer Stands

D. BlNKLEY AND P. SOLLINS

Soil Sci. Soc. Am. J. 54:1427-1433 (September-October)

Some values in Table 3 were 10 times too high; for Cascade Head, the annual rate of depletion of K + Ca + Mgshould be 0.54 kmoiyha, not 5.4 kmol^ha. Table 3 should be:

Table 3. Total and exchangeable (Exch.) soil element pools at the O- to 0.9-m depth.Ca

Location, stand

Wind RiverConiferAlder-coniferCascade HeadConiferAlder-conifer

Total Ckmol/ha

64908825

1949521580

Total

21283108

402439

Exch.

64.163.6

18.412.6

MgTotal

38403920

28002965

Exch.

45.848.1

37.714.7

KTotal

8951040

740797

Exch.

16.313.6

4.14.5

AlExch.

515505

353321